Elevator Pitch
DevOps isn’t just writing code; it’s fixing it. I use AI to simulate this by having it sabotage my projects. This forces me to practice root cause analysis on ‘live’ bugs, bridging the gap between student and Site Reliability Engineer.
Description
You can’t learn firefighting without a fire. Most DevOps education focuses on building (Day 1), ignoring the complexity of operating (Day 2). This talk introduces a novel approach: using Generative AI to simulate production chaos. We will explore how to prompt LLMs to act as adversaries injecting latency, breaking configs, and creating “ghost” bugs to force real-world debugging scenarios. By simulating the feedback loops of a live environment, we can master the “Three Ways” of DevOps without a job title. Don’t just learn to code; learn to survive the crash.
Attendee Takeaways:
Simulated Ops: How to create a “Virtual
Production Environment” locally.
Chaos Engineering: Using AI to inject realistic failures into Docker/K8s.
Continuous Learning: Automating the “Senior
Engineer” feedback loop for rapid growth
Notes
This talk bridges the gap between the theoretical ‘Three Ways’ of DevOps and the practical reality of modern AI tools. It moves beyond the standard ‘AI wrote my code’ narrative and offers a unique, pedagogical approach to learning Operations. It is highly visual, includes live prompt examples, and is designed to empower students and career-switchers to take control of their own learning curve.